This chapter is a report of the findings for the present study. Specifically, this study sought
understanding of the effects of an intervention including in-class coaching on teachers and children. As stated in Chapter 1, the first two research questions focus on the analysis of quantitative data to identify differences
between those who participated in an intervention including in-class coaching and those who did not. Answering the third question required qualitative methods to delve more deeply into the participating teachers' views of their intervention experience. This chapter is organized into three sections, each reporting the findings related to one of the three research questions. The specific research questions are as follows:
1. Did an intervention including coaching of
quality in preschool classrooms as measured by Early Childhood Environmental Rating Scale - Revised (ECERS- R)? If so, which environmental elements were most improved?
2. Did centers receiving an intervention including coaching for varying amounts of time differ in levels of school readiness?
3. What were prekindergarten teachers’ perceptions of coaching?
Research Question 1
The first research question was designed to examine the effects coaching may have on the quality of the
preschool classroom environments as measured by the Early Childhood Environmental Rating Scale - Revised (ECERS-R). Data from the Jacksonville Journey project were used. A paired-samples t-test was conducted using the pre-score, which was obtained before the implementation of coaching, and the post-score, which was obtained after a year of coaching. The data analysis addressed the following null hypothesis: There will be no statistically significant (p = .05) difference between the pre-score and post-scores in classrooms participating in an intervention including
coaching as measured by Early Childhood Environmental Rating Scale - Revised (ECERS-R).
Of the 30 centers assessed, many were eliminated from the study due to teacher turnover. Only centers with at least one consistent teacher in each assessment period were used. Nineteen classrooms in 19 centers were included. All of the classrooms were preschool classrooms. The classroom sizes varied from 8 to 19 children.
The alpha coefficient for the items used from the ECERS-R in the present study was .90, suggesting that scores on the items have a high internal consistency (Kline, 1999). The inter-subscale correlation showed
little to no correlation between items, indicating that the items were measuring different constructs, with little
overlap (Kline, 1999).
First, the overall scores were compared. There was a statistically significant difference in the ECERS-R overall pretest and posttest scores. A two-tailed paired samples t- test revealed that pre-scores (M = 3.586, SD = .906) and the post-scores (M = 4.1626, SD = .494), differed to a statistically significant degree, t(18) = 2.618, p =.008,95% CI[1.03, .11]. The p value was <.05. The confidence interval for the difference between the means allows for 95% certainty that the true value lies between 1.03 and .11. The posttest score was statistically
significantly higher than the pretest score; therefore, the null hypothesis was rejected.
Next, further testing was done using the scores of each subscale to identify which elements of the environment may particularly be affected by an intervention that
included coaching. A paired t-test was also conducted on the pretest and posttest scores of each of the six
subscales. It is important to note that the ECER-S has a total of 43 sub-items that can be used. However, in Duval County, only the first six subscales are used, representing 36 sub-items to assess classrooms. These scores are
presented in Table 1. Table 1
ECERS-R Pre and Post-Score Means Per Subscale (n=19)
Pretest Posttest Variable M SD M SD t p Cohen’s d Space and Furnishings 2.89 .58 3.56 .54 3.36* .001 .51 Personal Routines 2.35 .79 2.65 .64 1.36 .08 .20 Language and Reasoning 4.51 1.34 5.03 .85 1.51 .07 .22 Activities 3.95 1.0 5.21 .73 5.48* .00 .58 Interactions 4.31 1.76 3.83 1.07 .99 .16 .16 Program Structure 4.47 1.62 4.49 1.39 .02 .49 .01 Note. * = p < .05
The first subscale is Space and Furnishings. The items represented in this subscale include, indoor space, furniture for routine care, play and learning, furnishings for relaxation and comfort, room arrangement for play, space for privacy, child-related display, space for gross motor play, and gross motor equipment. The post-score was higher than the pre-score, and the difference was
statistically significant (t(18) = 3.36, p < .001, 95% CI [1.09, .25]). The effect size was measured using Cohen’s d, which is an appropriate test to measure the effect size between two means (Cohen, 1988). It indicates the
standardized difference between two means, and expresses this difference in standard deviation units. The effect size for the difference in means between the pre-score and post-score was .51 for the overall ECERS-R measure, which is considered a large effect size (Cohen, 1988
The second subscale is Personal Care Routines. This subscale is comprised of items including greeting and departing, meals and snacks, nap and rest, toileting and diapering, health practices and safety practices. This subscale did not show any statistically significant changes in the personal care routines present in the classroom from pre-score to post-score.
The third subscale is Language Reasoning. This
subscale includes books and pictures, encouraging children to communicate, and using language to develop reasoning skills and informal use of language. This subscale did not show any statistically significant change from pre-score to post-score.
The fourth subscale is Activities. This subscale is comprised of items related to the activities offered to the children. There was statistically significant growth
between the pre-score and post-score in this subscale,
t(18) = 5.48, p < .001, 95% [1.74, .77], with a large
effect size (d = .58) according to Cohen (1988). The items in this subscale include fine motor, art, music and
movement, blocks, sand and water, dramatic play, nature and science, math and numbers, the use of television and
computers, and promoting acceptance of diversity. The fifth subscale is Interactions. This subscale includes items related to supervision of gross motor activities, general supervision of children, discipline, staff-child
interactions, and interactions among children. This subscale showed no statistically significant change from pre-score to post-score.
The final group is Program Structure. This subscale includes items related to schedule, free play, and group
time. This subscale showed no significant change from pre- score to post-score.
The first research question was to determine if an intervention including coaching of prekindergarten teachers affect the environmental quality in preschool classrooms as measured by Early Childhood Environmental Rating Scale - Revised (ECERS-R). The results indicated that that the intervention including coaching may have led to an increase in items related to the learning environment of the
children and the activities available to them. However, statistically significant change was not obtained in the subscales related to interactions, language and reasoning, personal care routines, and program structure.
Research Question 2
In order to answer the second question, data were obtained from the Florida Department of Education. Data were obtained from all 361 centers that served children in Voluntary Prekindergarten in Duval County during the 2010- 2011 school year. Of these centers, Head Start centers and the Child Development Resource Center were eliminated due to the likelihood that the teachers of these centers
received additional coaching beyond that offered through the programs identified previously. After eliminating those centers, 344 centers remained. The centers were
sorted into three groups, including centers where teachers received no coaching in the two years previous to the
testing (N = 226), centers where teachers participated in one year of coaching directly preceding the testing (N = 78), and centers where teachers participated in two years of coaching directly preceding the testing (N = 40). The data analysis addressed the following null hypothesis: There will be no statistically significant (p = .05)
difference among centers participating in an intervention including in-class coaching for one year, those
participating for two years, and those that have not participated in an intervention including in-class
coaching, as measured by the Florida Kindergarten Readiness Screener (FLKRS).
An ANOVA was conducted to determine if there were differences among the three groups. The ANOVA results indicated that there were no statistically significant differences among the three groups, F(2,341) = .73, p = .48. Levene's rest for equality of variances was
calculated to determine if the variances of the groups were similar. However, the Levene's Test for Equality of
Variances was statistically significant (p = .041) which violates the assumptions of an ANOVA. Therefore the
samples having possibly unequal variances (Sawilowsky, 2002; Welch, 1951). The Welch test was not significant (p = .535), indicating that there were not statistically
significant differences among the three groups. Therefore, the evidence did not support rejecting the null hypothesis.
Research Question 3
The third research question focused upon teachers’ perspectives of in-class coaching. The semi-structured interview (Patton, 2002) served as the best means to collect descriptive data, using both structured and
flexible questioning. All individuals who volunteered to be interviewed, as a result of electronic invitation, were welcomed to participate and to share their experiences, opinions, and insights related to in-class coaching. Twelve participants, from four different preschool facilities, volunteered to partake in an interview, with each interview taking place at the preschool center in which the
participant was employed. Teachers experience varied from 2 to 20 years of working with preschool aged children. Only teachers who had worked with an in-class coach for 2 or more years were interviewed. The number of years each teacher worked with an in-class coach varied from 2 to 8 years. Of the 23 teachers, 3 spoke Spanish as their first language. The teachers varied greatly in age. Pseudonyms
were used to identify each teacher. Centers were labeled with a letter to avoid identification. Table 2 shows information about the 12 participants.
Probes and follow-up questions were used frequently to obtain more detail or additional information. My knowledge and experience in the field helped in both constructing the open-ended interview questions and analyzing the
participants’ responses. My own professional knowledge helped me to develop good follow up questions to clarify the participants’ explanations (Marshall & Rossman, 2011). However, my expertise also made it critical that I allowed the participants the autonomy and opportunity to express their own perspectives without my influence.
Each interview took place in the early afternoon, by request of each participant. Center directors monitored the participants’ classrooms while the interviewees and I stepped into a private location at the center. The first two center directors requested that interviews take place in the empty recreational space, and the final two
directors requested that the interviews take place in a break lounge. All other interviewees were encouraged to choose the place that they would find most comfortable. Each interview began with natural small talk concerning the weather and upcoming center events. Introductions took
place casually, including an exchange of names, a review of the purpose of the interview, and a review of the consent form. As I gained experience as an interviewer, my
strategies improved in regards to being able to prompt additional information based on the response to the
original questions, resulting in longer and more insightful interviews taking place last. Interviewees were given
opportunity to express their perceptions without
interruption or influence other than an occasional smile or nod to encourage continuing the conversation.
Table 2
Demographic Information of Participants
Pseudonym Center Native Language
Number of years working with an in-
class coach
Jennifer A Spanish 3
Janet A English 2
Jona A Spanish 2
Debbie B Spanish 5 (intermittently)
Donna B English 8 (intermittently)
Dana B English 6 (intermittently)
Carol C English 5 Catherine C English 3 Christen C English 3 Connie C English 5 Meagan D English 2 Mallory D English 2
Howe and Eisenhart (1990) argued that the process of data analysis should be made transparent in an effort to enhance the study’s credibility. The actual process of data analysis proved to be laborious and complex. When data collection was completed, transcription of recorded interview sessions began. This took several full days. I listened to each interview several times in order to ensure accurate transcription.
As Patton noted (2002, p.480), “Interpretation means attaching significance to what was found, making sense of findings, offering explanations, drawing conclusions, extrapolating lessons, making inferences, considering
meanings, and otherwise imposing order.” My knowledge and experiences as a former in-class coach and a professional in the field of early childhood education were beneficial in doing just this, though the process proved to be more tedious and difficult than I had imagined.
After I transcribed the interviews into text, the task of reading the transcript of each interview session began. I read the transcripts several times to gain a sense of the whole (Hatch, 2002) and to become intimate with the words of the participants (Patton, 2002). After reading the transcripts, I wrote memos in the margins to begin
stage of reflective memo writing is critical in order to move the analysis from the obvious to the unobvious. This was the case while completing this exercise. Ideas for codes began to become clearer.
Inductive analysis allows researchers to be immersed in the details of data to discover important themes and relationships (Johnson & Christensen, 2004; Patton, 2002). Hence, inductive analysis was employed in order to make sense and derive meaning from the voices of the
participants. In order to identify salient themes,
reoccurring ideas, and patterns of belief, the process of coding began, line by line (Strauss & Corbin, 1998). I began by highlighting with various colored markers words that could be grouped together under the same category or theme. The text was chunked into portions that were
closely related to one topic or theme. Sometimes the chunked text was as small as a word, while other passages continued for as long as a several sentences before the participant shifted focus. Often times the same chunk of text fit into more than one category. Each category was named using content-characteristic words (Strauss & Corbin, 1998). Color coding only lasted for the first several
pages of the transcript and then numbers were used to keep better track of emerging themes and categories, as the
limited color options were making it difficult to
differentiate between the many constructed categories. This process continued several times, as I refined
categories with each read through (Strauss & Corbin, 1998). Words that occurred frequently were seen as being
salient in the minds of respondents. D'Andrade (1995)
noted that "perhaps the simplest and most direct indication of schematic organization in naturalistic discourse is the repetition of associative linkages" (p. 287). When words, phrases, and ideas began to appear again and again, I began to group them and label them based on the concept they had in common.
Many themes emerged that were anticipated. These predictable themes are items listed in the mission
statements of the coaching project. They represent the fundamental processes of in-class coaching. The 10 themes are presented below, listed from those anticipated to those that were unexpected. The first four were those themes that were expected, as they are nearly identical to the definition of what coaches provide and the purpose of the coaching program. The four anticipated themes were the helpfulness of providing materials, assistance in setting up and planning for an appropriate classroom, providing
high-quality examples of instruction, and giving teachers suggestions to promote best practices.
The next four themes, validation, inspiration,
supporting English language learner teachers, and promoting developmentally appropriate practice, were themes
classified as being both unexpected and positive. The final two themes, lack of individualization and coaches’ level of competence, were classified as being unexpected and negative.
The first four themes were direct goals of the coaching program including the helpfulness of providing materials, receiving assistance in setting up and planning an appropriate classroom, receiving high quality examples of high quality instruction, and giving the teachers
suggestions to promote best practices. These themes were expected and are not expanded upon because they represent the basic goals of the coaching program and therefore were expected. They are inherently present because they define in-class coaching. In-depth explanations of the unexpected themes follow.
Of the 12 interviews, 5 of the teachers perceived the coaches to offer experiences that were categorized under the theme title “validation.” The teachers each offered tidbits that illuminated how their coaches somehow provided
them with a sense of validity as a teacher. Many of these teachers had never been observed or offered feedback on their strategies. For example, Janet, who worked with an in-class coach for two years, explained how her working with her in-class coach validated her.
When the coaches first came to me, I felt that there weren’t a lot of expectations with any of our
teaching. Nobody ever said how you were doing. So, when the coaches came in at first, I finally had positive feedback. . . . They were encouraging and told me how impressed they were. It meant a lot! Janet mentioned that she would often keep the written feedback forms that coaches would leave to reread on days she wasn’t feeling valuable.
This perception of coaches giving a sense of
validation was shared by several others including Christen. Christen’s face changed when we reached this point of the conversation:
I am now confident that I am a good teacher. I used to be so nervous for anyone to watch me. Now I get proud and excited. I used to be considered a
babysitter, and now I feel like a teacher. They
helped me know that I can teach with the best of them! Christen went from slightly reserved to beaming with pride while discussing her pride in her profession.
Another theme that emerged was the theme labeled “Inspiration.” Of the 12 teachers, 4 described instances related to how the in-class coach somehow stimulated the
desire to either return to school or pursue related credentials. Debbie recalled,
Honestly, it made me want to go back to school to
learn more. I had a class with ADD children, children with learning disabilities and some of the children had major problems. The coaches gave me resources from the county’s guidance office and helped me implement their suggestions. That motivated me to go further. I wanted to be a behavior specialist. . . . So I am back at school to get my bachelors [degree]. The coach was a major reason for that.
Debbie had completed more than half of her bachelor’s degree program at the time of the interview.
Mallory, a teacher with over 20 years of teaching, was inspired to go back to school to get her AA degree,
I was happy where I was. I didn’t even think that going back to school was an option. I never wanted to teach school age children, so I didn’t know a degree would be available to me in preschool education. I saw how young and excited my coach was, and I started to ask her about her degree. She encouraged me to go back to school. I did it very slowly because that is my style. But, I did it.
Mallory decided she would not continue her education beyond her AA degree because her center director stated that she would not be able to afford a pay increase to compensate her for her higher degree.
A theme that was salient across all 12 interviews was for items that I coded under the term “promoting
developmentally appropriate practice.” This theme was the label given for each teacher who referenced an example of
the coach helping her better understand what a child should